CD70 is a member of the tumor necrosis factor (TNF) family of cell membrane-bound and secreted molecules that are expressed by a variety of normal and malignant cell types. The primary amino acid (AA) sequence of CD70 predicts a transmembrane type II protein with its carboxyl terminus exposed to the outside of cells and its amino terminus found in the cytosolic side of the plasma membrane (Bowman et al., 1994, J Immunol 152:1756-61; Goodwin et al., 1993, Cell 73:447-56). Human CD70 is composed of a 20 AA cytoplasmic domain, an 18 AA transmembrane domain, and a 155 AA extracytoplasmic domain with two potential N-linked glycosylation sites (Bowman et al., supra; Goodwin et al., supra). Specific immunoprecipitation of radioisotope-labeled CD70-expressing cells by anti-CD70 antibodies yields polypeptides of 29 and 50 kDa (Goodwin et al., supra; Hintzen et al., 1994, J Immunol 152:1762-73). Based on its homology to TNF-alpha and TNF-beta, especially in structural strands C, D, H and I, a trimeric structure is predicted for CD70 (Petsch et al., 1995, Mol Immunol 32:761-72).
Original immunohistological studies revealed that CD70 is expressed on germinal center B cells and rare T cells in tonsils, skin, and gut (Hintzen et al., 1994, Int Immunol 6:477-80). Subsequently, CD70 was reported to be expressed on the cell surface of recently antigen-activated T and B lymphocytes, and its expression wanes after the removal of antigenic stimulation (Lens et al., 1996, Eur J Immunol 26:2964-71; Lens et al., 1997, Immunology 90:38-45). Within the lymphoid system, natural killer cells (Orengo et al., 1997, Clin Exp Immunol 107:608-13) and mouse mature peripheral dendritic cells (Akiba et al., 2000, J Exp Med 191:375-80) also express CD70. In non-lymphoid lineages, CD70 has been detected on thymic medullar epithelial cells (Hintzen et al., 1994, supra; Hishima et al., 2000, Am J Surg Pathol 24:742-46).
In addition to expression on normal cells, CD70 expression has been reported in different types of cancers including lymphomas, carcinomas, and tumors of neural origin. In malignant B cells, 71% of diffuse large B-cell lymphomas, 33% of follicle center lymphomas, 25% of mantle lymphomas, and 50% of B-CLL have been reported to express CD70 (Lens et al., 1999, Br J Haematol 106:491-503). CD70 is frequently expressed together with other lymphoid activation markers on the malignant Hodgkin and Reed-Sternberg cells of Hodgkin's disease (Gruss and Kadin, 1996, Bailieres Clin Haematol 9:417-46). One report demonstrates CD70 expression on 88% (7 of 8 cases) of thymic carcinomas and 20% (1 of 5 cases) of atypical thymomas (Hishima et al., 2000, supra). The second type of carcinoma on which CD70 has been detected is nasopharyngeal carcinoma. One study reports the presence of CD70 on 80% (16 of 20 cases) of snap-frozen tumor biopsies obtained from undifferentiated nasopharyngeal carcinomas (Agathanggelou et al., 1995, Am J Path 147:1152-60). CD70 has also been detected on brain tumor cells, especially glioma cell lines, solid human gliomas, and meningiomas (Held-Feindt and Mentlein, 2002, Int J Cancer 98:352-56; Wischlusen et al., 2002, Can Res 62:2592-99).
It has been proposed that transforming viruses including the Epstein-Barr virus (EBV) and the human T leukemia virus-1 (HTLV-1) can induce CD70 on cells such as epithelial cells that normally do not express CD70 (Agathanggelou et al., supra; Stein et al., 1989, Oxford University Press, p. 446). Therefore, expression of CD70 on malignant B cells could be a reflection of oncogenic transformation (Lens et al., 1999, supra). Also, since CD70 expression is induced on B cells after antigen encounter (Maurer et al., 1990, Eur J Immunol 20:2679-84; Lens et al., 1996, supra), stable expression of CD70 might reflect prolonged antigenic stimulation. This has been postulated to occur in follicular non-Hodgkin's lymphomas based on ongoing somatic hypermutation (Bahler et al., 1992, Proc Natl Acad Sci USA 89:6770-74; Bahler et al., 1992, Cancer Res 52:suppl. 5547S-51S).
The receptor for CD70 is CD27, a glycosylated type I transmembrane protein of about 55 kDa (Goodwin et al., 1993, Cell 73:447-56; Hintzen et al., 1994, supra). CD70 is sometimes referred to as CD27L. CD27, which exists as a homodimer on the cell surface (Gravestein et al., 1993, Eur J Immunol 23:943-50), is a member of the TNF receptor superfamily as defined by cysteine-rich repeats of about 40 amino acids in the extracellular domain (Smith et al., 1990, Science 248:1019-23; Locksley et al., 2001, Cell 104:487-501). CD27 is typically expressed by thymocytes, NK, T, and B cells (Hintzen et al., 1994, Immunol Today 15:307-11; Lens et al., 1998, Semin Immunol 10:491-99). On resting T cells, CD27 is constitutively expressed, yet antigenic triggering further upregulates CD27 expression (de Jong et al., 1991, J Immunol 146:2488-94; Hintzen et al., 1993, J Immunol. 151:2426-35). Further, triggering of T cells via their T cell antigen receptor complex alone or in combination with the accessory molecule CD28 releases soluble CD27 from activated T cells (Hintzen et al., 1991, J Immunol 147:29-35). Naïve B cells do not express CD27, but its expression is induced and, in contrast to CD70, sustained after antigenic triggering of B cells (Jacquot S et al., 1997 J Immunol 159:2652-57; Kobata T et al., 1995, Proc Natl Acad Sci USA 92:11249-53).
In marked contrast to the restricted expression of CD27 and CD70 in normal B lineage cells, both CD27 and CD70 are frequently co-expressed in many B cell non-Hodgkin's lymphomas and leukemias. This could potentially lead to functional CD27-CD70 interactions on these cells in the form of an autocrine loop, resulting in CD27 signaling and in CD70-induced proliferation, thereby providing a growth advantage to malignant cells (Lens et al., 1999, supra).
The available data supports a model in which ligation of CD27 on activated lymphocytes by CD70 delivers signals to the CD27-expressing cells, including co-stimulatory signals in T, B, and NK cells. (See, e.g., Goodwin et al., supra; Hintzen et al., 1995, J Immunol 154:2612-23; Oshima et al., 1998, Int Immunol 10:517-26; Smith et al., supra; Van Lier et al., 1987, J Immunol 139:1589-96; Gravestein et al., 1995, Int Immunol 7:551-7; Tesselaar et al., 1997, J Immunol 159:4959-65; Jacquot et al., supra; Agematsu et al., 1998, Blood 91:173-80; Kobata et al., supra; Agematsu et al., 1997, Eur J Immunol 27:2073-79; Sugita et al., 1992, J Immunol 149:1199-1203; Orengo et al., 1997, Clin Exp Immunol 107:608-13). Antibodies against both murine and human CD70 have been demonstrated to inhibit such activities, presumably by blocking the CD70/CD27 interaction (Hintzen et al., 1994, supra; Hintzen et al., 1995, supra; Oshima et al., supra).
Limited information is available on the modulation of cellular functions through CD70 signaling upon CD70/CD27 interaction, i.e., ‘reverse signaling’. Some CD70 antibodies have the ability to enhance T cell proliferation when they are presented to CD70-expressing T cells either cross-linked with a secondary antibody or immobilized on tissue culture plates (Bowman et al., 1994, J Immunol 152:1756-61; Brugnoni, 1997, Immunol Lett 55:99-104). Such ‘reverse signaling’ has also been described in a subset of B chronic lymphocytic leukemia (B-CLL) cells, and CD70 can function as a receptor to transduce signals to facilitate proliferation of PMA-stimulated purified B-CLL cells (Lens et al., 1999, supra). These observations suggest situations in which engagement of CD27 and CD70 can result in the delivery of agonistic signals to both the CD27 and CD70 expressing cells.
The role of CD70/CD27 co-stimulation in cell-mediated autoimmune diseases has been investigated in a model of experimental autoimmune encephalomyelitis (EAE) (Nakajima et al., 2000, J Neuroimmunol 109:188-96). In vivo administration of a particular anti-mouse CD70 mAb (clone FR-70) markedly suppressed the onset of EAE by inhibiting antigen-induced TNF-alpha production without affecting T cell priming, Ig production or TH1/TH2 cell balance. However, such treatment had little efficacy in established disease. It has been reported that expression of CD70 on T cells was enhanced by TNF-alpha and IL-12 and down regulated by IL4 (Lens et al., 1998, supra). Thus, the CD70/CD27 mediated T cell-T cell interactions may play a role in enhancing TH1-mediated immune responses rather than TH2-mediated responses. Supporting this hypothesis, the anti-mouse CD70 mAb FR-70 is also effective in inhibiting TH1-mediated collagen-induced arthritis (Nakajima et al., 2000, supra). In contrast, the same anti-mouse CD70 mAb did not show any efficacy in modulating lupus in NZB/NZW F1 mice and experimental Leishmania major infection in susceptible BALB/c mice, both of which are predominantly TH2-mediated autoimmue response (Nakajima et al., 1997, J Immunol 158, 1466-72; Akiba et al., 2000, J Exp Med 191:375-380).
The role of CD70 has not yet been investigated in acute graft versus host disease (aGVHD), another TH1-mediated immune response. GVHD is a major and often lethal consequence of allogeneic bone marrow transplantation (BMT) therapy that occurs when histocompatibility antigen differences between the BM donor and the recipient of the transplant are present (den Haan et al., 1995, Science 268:1476). GVHD is caused by mature T cells present in the transplanted marrow, as well as other minor cell populations (Giralt and Champlin, 1994, Blood 84:3603). It is noteworthy that CD70 has been detected in vivo on CD4+ cells in conditions characterized by allogeneic reaction, as in cases of maternal T cell engraftment in severe combined immune deficiency patients (Brugnoni et al., Immunol Lett 55:99-104). Prophylaxis of GVHD is achieved by pan-T cell immunosuppressive agents such as cyclosporine, corticosteroids, or methotrexate. In addition to the lack of specificity, these agents are also associated with significant adverse side effects. To limit these undesirable effects and the disruption of normal T cell functions, other therapeutic interventions based on selective targeting of T cells directly participating in allo-recognition and graft rejection are much needed.
CD70 is a potentially useful target for antibody-directed immunotherapy. As indicated supra, CD70 has a restricted expression pattern in normal cells: CD70 expression is mostly restricted to recently antigen-activated T and B cells under physiological conditions, and its expression is down-regulated when antigenic stimulation ceases. The key role of CD70 is believed to be facilitating plasma cell differentiation and contributing to the generation and maintenance of long-term T cell memory. Further, evidence from animal models suggests that unregulated CD70/CD27 interaction may contribute to immunological disorders, and, in humans, experimental data have also pointed towards potential abnormal regulation of the CD70/CD27 pathway in TH1-mediated immune disorders such as, e.g., rheumatoid arthritis, psoriasis, and multiple sclerosis. It is of particular interest that CD70 is expressed on a variety of transformed cells including lymphoma B cells, Hodgkin and Reed-Sternberg cells, malignant cells of neural origin, and a number of carcinomas.
Several groups have demonstrated the inhibitory effect of anti-CD70 mAb in both in vitro models of lymphocyte activation and animal models of TH1-mediated responses. The focus has been on the use of antibodies to block the CD70/CD27 co-stimulation pathway to achieve therapeutic efficacy. However, one main shortcoming of such an approach is the large number of signaling receptors, e.g., the CD28/CD80/CD86 co-stimulatory pathway, known to participate in immunological diseases. Consequently, blocking one specific signaling pathway may only have minimal impact on disease development. This is supported by the observations that anti-CD70 mAb can only partially inhibit in vitro T cell activation induced by allogeneic stimulator cells (Hintzen et al., 1995, supra) and an anti-CD70 mAb showed no therapeutic efficacy in EAE once the disease is established (Nakajima et al., 2000, supra).
Thus, there is a need in the art for developing an approach for depleting or inhibiting the growth of CD70-expressing cells involved in cancers and/or immunological diseases by means other than or in addition to blocking the CD70/CD27 interaction. As CD70 is expressed on the surface of mature antigen presenting dendritic cells, activated T cells, and activated B cells, agents that can target and inhibit or deplete CD70+ cells may prove to be effective in the removal of antigen presenting cells presenting autoantigens and offending autoreactive activated T or B cells, as wells as CD70-expressing tumor cells.
Approaches that have been used for increasing the therapeutic efficacy of antibodies are radiolabeling and combination with chemotherapy; however, these approaches include associated with undesirable side effects. For example, isotope therapy is associated with myelosuppression (Witzig, 2001, Cancer Chemother Pharmacol 48 (Suppl 1):S91-5), and combining therapy with antibodies and chemotherapeutics is associated with immunosuppression. Further, isotopically labeled substances are difficult to produce, and patients often experience relapse after initial treatment with isotopically labeled substances.
Accordingly, there is a need for anti-CD70 ADCs that are constructed in such a manner so as to be capable exerting a clinically useful cytotoxic, cytostatic, or immunosuppressive effect on CD70-expressing cells, particularly without exerting undesirable effects on non-CD70-expressing cells. Such compounds would be useful therapeutic agents against cancers that express CD70 or immune disorders that are mediated by CD70-expressing cells.